III Tax Structure
- Vito Tanzi, M. Yücelik, Peter Griffith, and Carlos Aguirre
- Published Date:
- October 1981
Apart from the level of taxation, it is important to consider the structure of the tax system. As already indicated, the structure of the tax system is the result of many different forces: (1) the characteristics of the economy, which to a large extent determine the availability of certain tax bases; (2) the expenditure programs of the government, which determine the extent of use of those tax bases; and (3) the political inclination of the government, which may prefer some types of taxes over others for equity or other reasons.
In general, developing countries have relied mainly on foreign trade taxes because they are easier to collect and often can be more easily justified in terms of some national objective, such as maintaining domestic employment or stimulating domestic industry. Developing countries have often relied on indirect domestic taxation, and particularly on excises, as the structure of the economy often prevented the administration of a full-fledged general sales tax. Income taxes have often been less important than either taxes on international trade or taxes on goods and services, and income taxes have often meant taxes on only a few large corporations and their employees and on government employees.
Table 7 analyzes the tax revenues of the countries of the sub-Saharan region expressed as ratios to GDP in order to permit an easier intercountry comparison. Data are given for a beginning year and for an ending year. Usually, the beginning year is in the early 1970s while the ending year is 1977 or 1978. The table indicates that import duties account for the largest part of revenues. For five of the countries listed (Botswana, The Gambia, Lesotho, Somalia, and Swaziland), import duties accounted for more than 10 per cent of GDP, which is a very high ratio; for several others, they accounted for more than 8 per cent of GDP. In fact, there were relatively few countries where these taxes accounted for less than 3 per cent of GDP—including Ethiopia, Ghana, Kenya, Tanzania, Uganda, and Zambia.
Table 7 also indicates that there is no clear-cut evidence that these ratios were lower in the ending year than in the beginning year.4 Therefore, unlike the conclusions reached by the literature on the change in tax structure that accompanies economic development, no indication of a movement away from import duties is evident, at least as long as the share of such duties in GDP is emphasized.5Table 7 also shows that export duties were quite significant, at least in a few countries. For example, export duties accounted for over 7.00 per cent of GDP in Uganda in 1978, 4.41 per cent of GDP in Rwanda in 1977, 3.27 per cent in Zaire in 1977, 3.14 per cent in Sierra Leone in 1978, and 3.03 per cent in Swaziland in 1977. In many countries this percentage was at least 2. The high share of export duties in GDP may be considered particularly disturbing in view of the negative effects that these taxes often have on production, allocation of resources, and exports. On the other hand, it can be argued that export duties, perhaps to a large extent, were considered substitutes for income taxes and, in some cases, were levied for short periods to prevent exporters from obtaining unusually high profits.
Although far more evidence is needed before any definite conclusion can be reached, questions can be asked (1) whether taxes on international trade have not been fully exploited by these countries, and (2) whether it would be wise to expect (or to recommend) substantial additional revenues from this source.
Taxes on goods and services range from relatively low amounts in countries such as Botswana, The Gambia, Nigeria, and Swaziland, where they account for less than 1 per cent of GDP, to very high ratios in countries such as Zambia, Somalia, Tanzania, the Congo, and Kenya, where they account for over 6 per cent of GDP. As expected, excises were generally more important than turnover or general sales taxes, although the latter were significant in a few countries such as the Congo, Kenya, and Tanzania.
With respect to taxes on income and profits, a very substantial spread is seen in total revenues from these sources. In comparison with taxes on international trade and even with domestic indirect taxes, taxes on income and profits are relatively unimportant, except where they include royalties from mineral exports. However, in a few countries, including Botswana, Liberia, Nigeria, Sudan, Togo, Swaziland, and Zambia, income taxes accounted for a sizable share of GDP (8 per cent or higher). From the statistical information available, it was not possible to separate clearly the relative importance of income taxes paid by individuals and those paid by enterprises. However, it seems safe to conclude that, in general, the share of income taxes in GDP was high only in those countries where the taxation of profits was substantial. From the limited evidence available, it seems that a large share of the income taxes on individuals was paid by government employees.
Table 8 provides the same breakdown of taxes, but the denominator is total tax revenue rather than GDP. Therefore, this table deals exclusively with the tax structure. The predominance of foreign trade taxes is even more pronounced than in Table 7. In eight countries—Benin, Botswana, Chad, the Comoros, The Gambia, Lesotho, Somalia, and Swaziland—import duties accounted for more than 50 per cent of total tax revenue. In 15 countries this proportion exceeded 40 per cent, and in 24 countries it exceeded 30 per cent. In a few countries (Benin, The Gambia, and Lesotho), the share of import duties in total tax revenues exceeded 60 per cent. These proportions are exceedingly high, even in the context of developing countries, where foreign trade taxes account, on average, for less than 40 per cent of total tax revenue. Such high proportions must inevitably raise the question whether the countries have given sufficient attention to developing domestic sources of revenue.
It is often found that unusually high proportions of total tax revenue are also derived from export duties. In a few countries (Burundi, Ethiopia, Ghana, Rwanda, Sao Tomé and Principe, and Uganda), export duties accounted for at least 20 per cent of total tax revenue. Therefore, if revenues derived from import duties are added to those from export duties, it is found that a rather surprisingly large share of total taxes derive from foreign trade. There is no doubt that the African countries of the sub-Saharan region have vigorously exploited the “tax handle” provided by the openness of their economies. The reliance on this tax base has made it possible for them to raise the total level of taxation substantially, but they may have paid a significant price in terms of resource allocation and efficiency.
Taxes on goods and services, although less significant than foreign trade taxes, were still relatively important6 and, in general, more important than taxes on income. For the majority of countries, taxes on goods and services generated revenue ranging between 10 and 30 per cent of the total. A few countries (Ghana, Kenya, Malawi, Niger, Sudan, Tanzania, and Zambia) had shares exceeding 30 per cent, while a few others (Botswana, The Gambia, Guinea, Nigeria, and Swaziland) had shares lower than 10 per cent. These proportions must be compared with those derived from income taxes, which generally accounted for less than, and in many cases substantially less than, 30 per cent of tax revenue. In a few countries (Botswana, Kenya, Malawi, Nigeria, Togo, and Zambia), however, income taxes provided more than 40 per cent of total tax revenue.
Tables 9 and 10 relate import duties to import values and export duties to export values, respectively, for the 1972–78 period. In other words, the foreign trade taxes are related more directly to their own tax bases. Table 9 shows that the average effective import duty for the period ranged from a low of 6.21 per cent for Zambia to a high of 35.92 per cent for the Central African Republic. For several countries (the Central African Republic, Ethiopia, Gabon, Ivory Coast, Somalia, and Sudan), this percentage exceeded 25 per cent; for several others (Benin, Burundi, Cameroon, Chad, the Congo, Madagascar, Upper Volta, and Zaire), it was between 20 and 25 per cent. For a few countries (Malawi, Tanzania, Togo, and Zambia), this percentage was lower than 10 per cent. The data in Table 9 indicate that the ratios of import duties to imports fluctuated considerably over the period, but no general trend is clearly evident. These ratios were relatively high and were the result of high tariffs, combined with exemptions that were very significant in some countries (Sudan and Somalia). Table 9 should be studied together with Table 11, which gives information about the level of statutory duties in these countries.7
Table 10 shows the share of export duties to export values, ranging from very low values for some countries (Nigeria, Botswana, Seychelles, and Senegal) to values exceeding 20 per cent in other countries (Burundi, Ghana, Rwanda, Uganda, and Zaire). Different views exist about the appropriateness of using export taxes; however, taxes of this magnitude are likely to have disincentive effects on production and exports, especially if they are not temporary. What is more disturbing is the apparent upward trend in the ratios in several countries, shown in Table 10. However, to the extent that high export taxes are levied for limited periods, either to reduce the profits of exporters during devaluations or to reduce their gains during export booms associated with sharply higher export prices (for example, the coffee boom), their disincentive effects are likely to be limited while their stabilization effects may be substantial.
Table 15, which gives further details on export duties, shows the tax base, the main items that are exported, whether the rates are specific or ad valorem ones, and the rate of the export duty. For this table, too, a word of caution is necessary. As it has been assembled from several sources, the information for some countries is more recent than for others.
An attempt to collect comparative data on the legal structure of income taxes levied on individuals was unsuccessful because of the complexity and the diversity of systems of personal income taxation. It was difficult to take into account in tabular form the treatment of the family, the schedular nature of many taxes, and the different methods for granting personal exemptions. In general, it could be concluded that the tax rates were relatively low on incomes of, say, up to 6 to 8 times the per capita income of the countries, and these rates became very high on incomes of only a few taxpayers (incomes over 20 times the per capita income of the country). Table 12, on the other hand, yields some comparative information on the legal treatment of company income. The standard rate on company income is generally between 35 per cent and 50 per cent, which is normal by international standards. In a few countries (Ghana, Mauritius, Sudan, and Zaire), however, the standard rate may go up to 60 per cent, which seems high by international standards.
Tables 13 and 14 add information on domestic indirect taxes. Table 13 relates to general sales taxes, showing that a large number of the countries of the region have some kind of general sales tax, while Table 14 relates to excises. In reality, these taxes are likely to be far less general than is assumed and are likely to collect much of the revenue at the point of entry of imports. Nevertheless, it is surprising that so many countries have moved beyond excises in order to extend their indirect taxes. In most of the countries, the basic rate of these sales taxes ranged between 5 per cent and 15 per cent. In several countries the sales tax was levied at a single rate on products and a different rate on some services. In other countries the sales tax itself differentiated among products by taxing them at various rates. In specific countries (for example, Lesotho and Senegal), the spread in the rates was substantial.
Table 14 attempts to give an idea of the fiscal importance of products (petroleum, beer, and tobacco) that contribute mainly to revenue from excises. It shows that, for the countries for which this information could be gathered, the contribution of petroleum products, expressed as a percentage of GDP, exceeded 1 in two countries (Senegal and Zambia) and exceeded 0.50 per cent in several others. Beer, on the other hand, generated 1.62 per cent of GDP in Rwanda and 3.60 per cent in Zambia. Cigarettes contributed more than 1 per cent in Sierra Leone. Table 14 also shows that in the majority of the countries for which data are presented the contribution of excises has been declining over recent years. This drop was undoubtedly due to the specific nature of these taxes, combined with inflationary situations. Table 16 shows the extent to which excises were, in fact, specific.
Many countries of sub-Saharan Africa have increased their customs duty revenue by gradually raising the rates of customs duties, especially on items of conspicuous consumption and industrial imports in particular (on equity and effective protection grounds). In addition, the Lomé Convention, signed on February 28, 1975, also had a favorable impact on customs duty revenues. Under this Convention, the European Community offered complete tax exemption for imports from 46 developing countries of Africa, the Caribbean, and the Pacific without requiring reciprocity from them. This encouraged the sub-Saharan African countries, which were signatories to the Convention, to remove customs duty preferences they had granted to the imports from some or all Community member countries in the past and, as a result, these countries earned larger customs duty revenues.
However, the period covered is perhaps too short to test the conclusions of what is essentially a long-run proposition.
The distinction between “taxes on goods and services” and “foreign trade taxes” is somewhat blurred in sub-Saharan Africa, as many countries tax luxury consumption but categorize the receipts as import duty receipts, and many others levy sales taxes but collect these taxes primarily on imports.
It should be realized that the ratios in Table 9 are not necessarily directly related to the level of nominal import duties, as very high duties on some products, combined with exemptions on others, could distort the structure of imports to such an extent as to theoretically reduce to zero the ratio of import duties to imports.